Droughts have catastrophic impacts across a wide range of sectors. Proper assessment of drought is central to water management; early identification can help mitigate the negative impacts of a drought. A number of metrics have been developed towards this end and are currently in use for characterizing droughts. One major challenge in computing these indices, however, is the integrity of the original datasets. Whereas precipitation is not significantly affected by anthropogenic changes, if we discount climate change, other local variables, such as soil moisture and streamflow, are greatly influenced by land use land cover changes. Streamflow, which is the most important parameter in water management, do incorporates meteorological forcings, even though not as first order response but filtered by watershed characteristics, but is significantly affected by dams, diversions, return flows, reduction of base flows by excessive groundwater pumping, and urbanization. In this study we use a land surface model (LSM) to generate runoff in the Rio Grande basin and compute the standardized runoff index. We test four copulas to assess which one is most suited to model the dataset and present preliminary results of the joint probabilities of drought severity and duration, along with the conditional probability distribution charts of drought severity given a threshold duration, and drought duration given a threshold drought severity.
Date of this Version